1. Cross-Modal Associative Mnemonic Signals in Crow Endbrain Neurons
Felix W. Moll, Andreas Nieder 
Current Biology 
Volume 25, Issue 16, Pages (August 2015)
DOI: /j.cub
Copyright © 2015 Elsevier Ltd Terms and Conditions

2. Current Biology 2015 25, 2196-2201DOI: (10.1016/j.cub.2015.07.013)
Copyright © 2015 Elsevier Ltd Terms and Conditions

3. Figure 1
Audio-Visual Association Task: Behavioral Protocol and Behavioral Performance
(A) The crows initiated a trial by moving their heads within the range of a light barrier. This was fed back by a visual go stimulus. Then an auditory sample was played, followed by the delay after which crows had to choose the correct associated color stimulus to gain a food reward. In 50% of the trials, the first test color was a match, whereas the first test stimulus was a non-match in the other 50% of the trials.
(B) Average behavioral performance per session and bird for all recording sessions.
(C) Average behavioral performance per association and bird across all recording sessions. Error bars indicate the SEM over sessions.
Current Biology  , DOI: ( /j.cub )
Copyright © 2015 Elsevier Ltd Terms and Conditions

4. Figure 2
Working Memory-Related Association-Selective Activity in NCL Neurons
(A–C) Example of an individual association-selective neuron preferring the noise-blue association.
(A) Dot raster showing the neuron’s response in individual trials, ordered by the presented auditory sample cue (correct trials only). Each dot signifies one action potential. Vertical lines mark transitions between pre-sample, sample, delay, and test period.
(B) Error trial dot raster in comparison to correct trial dot raster in (A).
(C) Peri-stimulus time histogram (PSTH), obtained by averaging the dot rasters and smoothing with a 150-ms boxcar window. Note that error trial curves are inverted compared to correct trial curves.
(D–F) Example neuron preferring the tit song-red association.
(D) Dot raster correct trials.
(E) Dot raster error trials.
(F) PSTH correct and error trials.
(G) Average normalized PSTH for the population of delay-selective neurons for their preferred and non-preferred association. Solid lines show the population response (n = 119) in correct trials and dotted lines (n = 71) in error trials. Shaded areas indicate SEM over neurons.
(H) Quality, temporal evolution, and latency of association selectivity for all delay-selective association neurons. Each line represents one neuron. Neurons are sorted by their association preference during delay and the latency of association selectivity. Each neuron’s latency is marked by the white line that runs across sample and delay period. White vertical bars mark transitions between task periods. Noise-blue-preferring cells are, by convention, represented by delay AUROC values < 0.5 (and vice versa for tit song-red-preferring cells).
Current Biology  , DOI: ( /j.cub )
Copyright © 2015 Elsevier Ltd Terms and Conditions

5. Figure 3 Error-Predictive, Internally Generated Pre-sample Activity
(A) Average pre-sample discharge rates of all delay-selective neurons for correct and error trials. Incorrect behavioral responses during preferred association trials were predicted by decreased pre-sample discharge rates compared to increased pre-sample discharge rates during incorrect non-preferred association trials.
(B) Histogram of pre-sample period AUROC values of all delay-selective association neurons (correct trials only).
(C) Same as (B), but for error trials. Note that this distribution approximates the distribution in Figure S2B.
(D) Average pre-sample AUROC values of all delay-selective neurons for correct and error trials. Noise-blue-preferring cells are, by convention, represented by delay AUROC values < 0.5. In the pre-sample phase of error trials these noise-blue cells had an average AUROC value > 0.5 (and vice versa for tit song-red-preferring cells), which predicted the upcoming incorrect behavioral response.
Error bars indicate the SEM over neurons.
Current Biology  , DOI: ( /j.cub )
Copyright © 2015 Elsevier Ltd Terms and Conditions

6. Figure 4
Working Memory-Related, Auditory Stimulus-Selective Activity in NCL Neurons
(A) Auditory match-to-sample task: the crows initiated a trial by moving their heads within the range of a light barrier. This was fed back by a visual go stimulus. Then an auditory sample was played, followed by the delay after which crows had to choose the correct auditory test stimulus to gain a food reward.
(B) Example of an individual neuron responding selectively to auditory stimuli during the delay phase. Top: dot raster showing the neuron’s response in individual trials, ordered by the presented auditory sample cue (correct trials only). Each dot signifies one action potential. Vertical lines mark transitions between pre-sample, sample, delay, and test period. Bottom: PSTH (correct trials only), obtained by averaging the dot raster and smoothing with a 250-ms boxcar window.
(C) Percentage of stimulus-selective NCL cells during sample and delay periods for the audio-visual task (gray bars) and for the auditory match-to-sample task (white bars).
Current Biology  , DOI: ( /j.cub )
Copyright © 2015 Elsevier Ltd Terms and Conditions